New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
As others have mentioned, if you cannot stand downtime, what you need is a high-availability setup. You might get downtime with the bigger hosts too.
For now it's back up so I'll see if it stays that way....> @neon_orange said:
Exactly what I meant
I don't use any providers on here, either.
Just here for the memes.
Someone else will say the opposite. They'll have outages at the big corporations and years of 100% uptime at the low end host. I mean, it's obvious. Remember the Office 365 outages of last year? Or when Linode was basically dead in the water for a long period that forced them to upgrade their networks that one time several years ago? I've had multiple outages at Hetzner and OVH. I've assisted in causing at least one outage at DigitalOcean (rate limiting internal scans for vulnerable software, to notify clients, is important).
That's the thing about dice rolls and random chance: it's random. No computer stays online forever.
Well that's because of my experiences that i'm aware that 100% is like looking for unicorns ^^. But i may have missed some providers able to provide 100% uptime and providing 100% SLA as well? Please share some names i would like to know!
Also if you knew SRE you would know that 100% service availability is against our goal ^^.
I am using hetzner and netcup for 2+ years without a downtime. That seems like 100% to me. Yeah sure. A bit of downtime after 3 years. That, I will still consider that 100%.
I think you are confused here. I don't think anyone (op or me) is talking about dedicated bare metals here. Third word in my initial sentence was "vps".
Please read about clustered virtualization.
EDIT: I meant 100% uptime in the perspective of a customer. Yeah shit can burn down any second when you work at a DC. But, service can be maintained 100% for virtual servers, unless extreme bad luck i.e. 1 outage in 3 years.
I think it makes a huge difference whether they properly update network status page with outages and solutions. More transparency allows you to see their track record.
The worst downtime was from Oracle at 8 days. My VPS was stuck in a shutdown state and couldn't be started/stopped. Fucking status page would downplay the severity and then say solution is being implemented, but like 5 days until my host gets that fix. It must have one of the very last to get it.
They really don't give a fuck (no sense of urgency) and just say to restore from backup. But they make it a manual backup process! Thankfully, they did a blog post that had a script to backup the boot volume everyday.
First time? Stay here for a little longer and you'll see what we called LET drama. It is entertaining, better than 9gag.
Anything is better than 9gag
sthosting? their products look good, but still considerably new as LE provider. look for more established ones who have a proven (or rather recorded) track record.
and always have remote backups...upcloud seems good for claimed 100% uptime.
AWS also got lots of downtimes recently. I don't think that we can rely on any single provider. It must be a high availability setup between multiple regions/providers. In the downtime scenario, we can just route the users to a healthy region and everything is good.
I got attracted to LET by the Racknerd deals last BF. At first, it looks too good for 4GB Ram 3 cores for $44/year (~$3.7/month). Then, I realize that the reason is that they use old hardware. So, I start to look for other options and the Hetzner US is great ($4.44/month - 2GB EPYC CPU). However, I see some discussions about limited DDOS protection there.
Then, I bought some VPS at @Hybula for their advertised 480Gbps Protection & good speed to the US (~$5.17/month - 2GB EPYC 7302P). After that, I got the @terrahost deals for VPS Ryzen 5950x (~$4.25/month) with GB5 1700. I do get some GreenCloudVPS Singapore VPS for the APAC region too.
According to this thread, LiteServer seems a good option so I will wait for their deal again.
And even that is just until your registrar screws up like enom did for days. Or when an upstream accepts a bad announcement for your IP space. That's why I always say you'll never truly have zero single points of failure, and you'll spend all of your money in pursuit of it. We just have to accept that everyone has outages. Facebook, Google, Microsoft, everyone. We're all in good company.
Many of the providers advertising in this forum are mediocre or shit, and will not be here in a few years from now.
Your ability to distinguish between fly-by-night operations and reliable providers will determine the outcome of your business here.
Anyway, if you want small companies to compete in pricing with Hetzner, your expectations are unreasonable to start with.
The thing is, whether low-end or high-end, I've seen outages happen on both. I am using a mix of low-end hosts for ONDDNS but because of the nature of the design, I can theoretically afford downtime because of how the data replicates to all servers. So I would say that I feel (this is just a personal opinion) that it's 99% on the design of what you're providing to the public. Obviously this gets complicated when you don't own the software you're working with or are unable to adjust it to suite your needs. Thankfully ONDDNS has been relatively simple but maybe my next step will be moving from just DNS serving to WEB / HTTP serving and that will require a little more finess BUT not alot so far (I've been thinking and designing it in my head)
Moral of the story - if you want uptime - never trust a single point nor a dual point. You need to think in 3's or higher.
AWS is the new IBM. Customers will be fine if you use aws and it is down(lots of other stuff will be down too). If you use something else, expectations will be different.
For me, @MaxKVM and NexusBytes @seriesn VPS are rock-steady. Almost zero downtime.
Luckily, I bought them at a discounted price. So, I am paying a low-end price for good VPS.
Good, you have high IQ thats why.
People like me settle for fiberia.io
Issues tend to happen, even with the "big" providers.
With low-end providers, it's best if you set your expectations at the low end as well, or none at all.
However, there are low-end providers who do have high-end passion and determination and would be more than willing to work with their customers.
No way aws bro. Week ass expensive hardware and bad customer support for small customers.
goodbye and thanks for all the fish.
Been using phpfriends since 2020, 2 downtimes for maintenance 15 mins each.
After 8 years of lurking around LET, I experienced the exact same thing. I have settled with Hetzner as well.
What pissed me off the most using the LET No. 1 provider was the lack of ability to scale up and down the VPS at will. I had to create a ticket, wait 12 hours (due to timezone differences) only to receive a 1 year upgrade up sell deal. I realized the dirt cheap VPS is not worth it at all. It's nice for hobby throwaway projects.
AWS:
https://hetrixtools.com/report/uptime/82353f457499aca3d3583245d63226ff/
Frequency=1 minute
1 downtime<1 minute in 4 years
But slack was down recently due to AWS. Good report shown here but that doesn't mean they didn't have downtime for other customers. That's the catch.
I have a problem, i.e., I'm addicted to collecting VPS. I used to spend about $200~400/mo on Google Cloud, AWS literally for nothing, until I found LET. That really saved me, and I haven't wasted that much ever since. Now, I have only 11 small boxes from the so-called low-end provider, idling, of course. I cannot say that I'm cured, but I do live a happier life
im happy with @dustinc
Just wanted to highlight this when I came across this report. Never seen a report this clean before. 4 years, 1 downtime<1 minute. Its really a spectacle.
So long?
It is bogus thinking when you think that you will dump lowend. It is lowend dumping you. And once you felt the rush, the tingles from bargain deal - lowend hooked you for life. It can't be quit cold turkey, without major mental restructurization.
Enormous amount of endorphins, which is result of buying cheap deal, is not to be ignored.